Mapping solar potential obstructions

using LiDAR dataKrista Amolins

David Coleman, Yun Zhang, Peter DareUniversity of New Brunswick, Fredericton, NB

ASPRS 2011

/20

Sunlight in Milwaukee

2

Sunrise, sunset, dawn and dusk times Sun path

Dec 21

Jun 21

May 1

Image Source: http://www.gaisma.com

/16

Insolation, in kWh/m2/day

3

0

2.0

4.0

6.0

8.0

J F M A M J J A S O N D

Fredericton Iqaluit Milwaukee

Source: NASA Langley Research Center Atmospheric Science Data Center, via http://www.gaisma.com

If could capture and convert 100% of the solar radiation hitting a 1 m2 area on May 3:- power 5 100 W light bulbs for 10 hours- make 25 pots of coffee- dry 1 load of laundry

BUT high efficiency solar panels convert only 18%

http://geonb.snb.ca/geonb/N

Sun angles December 21

rise 124°set 236°

max elev 20.7°

Limited Solar Potential

Fredericton, NB, Canada

Worst CaseBest Case:

Obstructions...

/20

LiDARpoint cloud

solid surface

worst case obstructions

solid object points

best case obstructions

Workflow for Mapping Obstructions

5

LiDARpoint cloud

Refinements...

/20

LiDAR Point Cloud

6

50 m

Data collected November 2007 ➙ leaf-offUp to four returns per pulse; most last returns from ground

/20

Solid Surface

750 m

Grid resolution 1 mSimple filtering: last returns.Contains ground, buildings, evergreen trees and hedges, cars, some artefacts from deciduous trees.

/20

• Worst case means: on the shortest day, when the sun does not reach a high elevation, will sunlight reach the surface?

• Compare LiDAR points to surface cells: are objects obstructing the sun at any time?

Worst Case Obstructions

8

/20

Worst Case Obstructions

9

LiDAR Point Cloud Solid Surface

/20

Worst Case Obstructions

9

LiDAR Point Cloud Solid Surface

Sun angles December 21

rise 124°set 236°

max elev 20.7°

/20

Worst Case Obstructions

9

LiDAR Point Cloud Solid Surface

Sun angles December 21

rise 124°set 236°

max elev 20.7°

/20

LiDARpoint cloud

solid surface

worst case obstructions

solid object points

best case obstructions

Workflow for Mapping Obstructions

10

solid object points

Refinements...

/20

Solid Object Points

11

Data collected May 2006 ➙ leaf-onPoints classified using intensity, local density, and local variation in height

/20

Solid Object Points

11

Non-solid Objects

Data collected May 2006 ➙ leaf-onPoints classified using intensity, local density, and local variation in height

/20

• Best case means: (on the shortest day)are any of the obstructions not solid objects?

• Compare only solid object LiDAR points to surface cells: are objects obstructing the sun at any time?

Best Case Obstructions

12

/20

Best Case Obstructions

13

Greyscale: Solid surfacePink: Non-solid Object Points

Sun angles December 21

rise 124°set 236°

max elev 20.7°

/20

Greyscale: Solid surfacePink: Non-solid Object Points

Best Case Obstructions

14

/20

Best Case Obstructions

14

Non-solid Objects Included

/20

Best Case Obstructions

14

Non-solid Objects Included Non-solid Objects Excluded

/20

LiDARpoint cloud

solid surface

worst case obstructions

solid object points

best case obstructions

Workflow for Mapping Obstructions

15

Refinements...

/20

Refinements - Directional

16

Greyscale: Solid surfaceGreen: Non-solid Object Points

East

South

West

/20

Refinements - Directional

16

Greyscale: Solid surfaceGreen: Non-solid Object Points

East

South

West

/20

Refinements - Directional

17

Greyscale: Solid surface

East

South

West

/20

Refinements - Directional

17

Greyscale: Solid surface

East

South

West

/20

Recap

18

• Renewable energy sources, e.g., solar

• Low insolation in winter

• Filter LiDAR data: ground (solid surface) and obstructions (everything)

• Classify LiDAR data: solid objects (not trees, power lines...)

• Analyze obstruction direction, seasonal solar variations to refine results

/20

Further Work• Apply better filtering methods

• Smooth solid surfaces

• Quantitative assessment of results

• Improve classification

• Separate trees from power lines

• Locate building edges

• Apply to vertical surfaces19

Thank You!